Critical questions for big data, Provocations for a cultural, technological, and scholarly phenomenon. Danah boyd&Kate Crawford
Big Data for Education: Data Mining, Data Analytics, and Web Dashboards. Darrell M.West
Critical questions for big data, Provocations for a cultural, technological, and scholarly phenomenon. Danah boyd&Kate Crawford
Overview:
This literature is a critical article. It mainly illustrates as the era of Big Data is underway, Big data is more and more popular rearch tool in many academic fields. Nevertheless, some problems occur as the rai of popularity of Big Data. This literature focus on the six main critical questions for Big Data.
Introduction:
The first part is the introduction. In this part, 随着大数据的普及，越来越多学术领域将大数据
作为一种新的工具。简单阐述了大数据在一些领域的用途后，将重心放在大数据被运用的同时，很多相关的问题滋生了。
Main Body:
1. Big data changes the definition of knowledge
In the first main critical question, author quoted the Fordism at the beginning. Fordism was a new t of tools to manufacture. Fordism meant automation and asmbly lines, when automation and asmbly lines became the main manufacturing method, skilled craftspeople and slow work was no more important. It means that Fordism change the manufacturing model. As same as Ford changed the way we made cars，big data is not only a tool that we changed, but also the changing of the older system of knowledge happens in the meanwhile. In addition, the author pointed out that other methods for ascertaining why people do things, write things, or make things are lost in the sheer volume of numbers when Big Data becomes popular. This is not a space that has been welcoming to older forms of intellectual craft. Last but not the least, before the Big Data b
ecomes the majority rearch method, the own inbuilt limitations of Big Data must be taken into account.
Comments:
The author quoted Fordism to state that when we change the tool, we also change the model and method. It is proper, we must consider the influence of the changing of system of knowledge. However the author didn’t point out the limitations of the Fordism. This makes less persuasion for the author’s point of limitations of Big Data must be taken into account, as we also cannot e the bad aspect of the changing of system of knowledge.
2. Claims to objectivity and accuracy are misleading
At the beginning of this part, author quoted the Latour’s view that in social domain, there is a line between what is and is not quantifiable knowledge. After that the author illustrated that in real world, working with big data is still subjective. As an example, when considering messages from social media, people filter the message subjectively, and usin
g the messages to produce facts. As the interpretation is at the center of data analysis, regardless of the size of a data, it is subject to limitation and bias. This will lead to the misinterpretation.
Comments: when we consider about a rearch, the subjective judgment is always there. Working with Big Data will enlarge this problem, as it can be applied in many objective reports. One thing that I suppo can be applied for this problem is reduce the number of process of Big Data, the less you do, the more objective results you have.
3. Bigger data are not always better data
Author stated Big Data prents us with large quantities of data, the understanding of sample is more important than before. To support this viewpoint, author quoted an example of Twitter, Twitter provides an example in the context of a statistical analysis. This example illustrates two critical point, the first point is Twitter does not reprent all people, and it is an error to assume people and Twitter urs are synonymous. Which also means that Big Data and whole data are not the same. The cond point is it is hard
to understand the sample when the source is uncertain. Twitter creates many subts of its all ur, which contains all public tweets, 10 percent of public tweets, 1 percent of public tweets respectively. It is not clear what tweets are included in the different data streams. Without knowing, it is difficult for rearchers to make claims about the quality of the data that they are analyzing. Thus when we combine multiple data ts, it is increasingly important to recognize the value of ‘small data’.
Comments:
Indeed, bigger is not always better for data, when the rearchers focus on the Big Data, the small data will be ignored. It is better to think both of Big data and small data. When we u Big Data in a rearch, it can provide a rough result, then we apply the small data to make the result more specific and preci.
4. Taken out of context, Big Data los its meaning
As large data ts can be modeled, data are often reduced to what can fit into a mathem
atical model. Taken out of context, data lo meaning and value. In this part, author refer the social network sites to identify personal connections. Big Data introduces two new popular types of social networks derived from the data traces ‘articulated networks’ and behavioral networks, which reprent two different groups. Both of the two have great value to rearchers, but they are not equivalent to personal networks. For example, not every connection is equivalent to every other connection, and neither does frequency of contact indicate strength of relationship. Data are not generic.
